Strategic Double Clicking - Page 132

S485122 · 2017-05-08, 04:37

Quote:

Originally Posted by LaurV

Here is one I am not happy with, M76453229. I ran both tests and the residues matched on the way, but one is clLucas and one is gpuOwl, both with null shift, in the same card (HD7970). Most probably (99.999999%, hehe) my residue is right, but a "shifted" check won't hurt.

On the status page it is set as verified, but only one LL line appears even if in the History both do...

Jacob

LaurV · 2017-05-08, 05:47

Yes, this is supposed to happen, it appears as verified, because two different programs were used, but James' parser is clever enough and identifies that only one shift value was used. I will not do a TC, because in that case a QC will still be needed (different user, blah blah). Somebody else may queue it. Thanks in advance.

Madpoo · 2017-05-09, 15:19

Here's a "top 100" list of potentially bad runs, looking at their monthly as well as total track records. I re-sorted afterwards by exponent to make things tidier (didn't do that last time...)

Code:

exponent	Bad	Good	Unk	Sus	Solo	Mis	worktodo
43998749	1	0	1	0	1	0	DoubleCheck=43998749,72,1
48170849	1	0	1	0	1	0	DoubleCheck=48170849,73,1
48196541	3	0	1	0	1	0	DoubleCheck=48196541,72,1
48454789	2	0	3	0	3	0	DoubleCheck=48454789,72,1
48572149	2	0	1	0	1	0	DoubleCheck=48572149,72,1
48631787	2	1	1	0	1	0	DoubleCheck=48631787,72,1
48696377	2	1	2	0	2	0	DoubleCheck=48696377,72,1
48797741	2	0	1	0	1	0	DoubleCheck=48797741,72,1
49003573	3	1	1	0	1	0	DoubleCheck=49003573,72,1
49396771	2	0	3	0	3	0	DoubleCheck=49396771,72,1
49592891	2	1	1	0	1	0	DoubleCheck=49592891,72,1
49719167	1	0	3	0	3	0	DoubleCheck=49719167,72,1
49827257	2	0	2	0	2	0	DoubleCheck=49827257,72,1
49975669	2	0	2	0	2	0	DoubleCheck=49975669,73,1
50177599	1	0	1	0	1	0	DoubleCheck=50177599,73,1
50194649	1	0	1	0	1	0	DoubleCheck=50194649,73,1
50206487	1	0	2	0	2	0	DoubleCheck=50206487,73,1
50206537	1	0	2	0	2	0	DoubleCheck=50206537,73,1
50229737	2	0	3	0	3	0	DoubleCheck=50229737,73,1
50366441	2	0	2	0	2	0	DoubleCheck=50366441,73,1
50451007	2	1	1	0	1	0	DoubleCheck=50451007,73,1
50453971	2	1	1	0	1	0	DoubleCheck=50453971,73,1
50456467	1	0	1	0	1	0	DoubleCheck=50456467,73,1
50465879	1	0	2	0	2	0	DoubleCheck=50465879,73,1
50575379	1	0	1	0	1	0	DoubleCheck=50575379,73,1
50601799	2	0	1	0	1	0	DoubleCheck=50601799,73,1
50608559	1	0	2	0	2	0	DoubleCheck=50608559,73,1
50658019	2	0	1	0	1	0	DoubleCheck=50658019,73,1
51578843	3	2	4	0	4	0	DoubleCheck=51578843,73,1
51578929	3	2	4	0	4	0	DoubleCheck=51578929,73,1
51579013	3	2	4	0	4	0	DoubleCheck=51579013,73,1
51579097	3	2	4	0	4	0	DoubleCheck=51579097,73,1
51751211	1	0	1	0	1	0	DoubleCheck=51751211,73,1
51984847	1	0	3	0	3	0	DoubleCheck=51984847,73,1
52065473	1	0	1	0	1	0	DoubleCheck=52065473,73,1
52094267	1	0	3	0	3	0	DoubleCheck=52094267,73,1
52139429	1	0	3	0	3	0	DoubleCheck=52139429,73,1
52139453	1	0	3	0	3	0	DoubleCheck=52139453,73,1
52139473	1	0	3	0	3	0	DoubleCheck=52139473,73,1
52362521	1	0	3	0	3	0	DoubleCheck=52362521,73,1
52380143	1	0	1	0	1	0	DoubleCheck=52380143,73,1
52383911	1	0	1	0	1	0	DoubleCheck=52383911,73,1
52422059	2	0	2	0	2	0	DoubleCheck=52422059,73,1
52541717	1	0	3	0	3	0	DoubleCheck=52541717,73,1
52683481	1	0	2	0	2	0	DoubleCheck=52683481,73,1
52717927	1	0	1	0	1	0	DoubleCheck=52717927,73,1
53212283	1	0	1	0	1	0	DoubleCheck=53212283,73,1
53487703	1	0	1	0	1	0	DoubleCheck=53487703,73,1
53647547	5	4	9	0	9	0	DoubleCheck=53647547,73,1
53648423	5	4	9	0	9	0	DoubleCheck=53648423,73,1
53648893	5	4	9	0	9	0	DoubleCheck=53648893,73,1
53648981	5	4	9	0	9	0	DoubleCheck=53648981,73,1
53889389	1	0	3	0	3	0	DoubleCheck=53889389,73,1
54096473	1	0	1	0	1	0	DoubleCheck=54096473,73,1
54169523	1	0	1	0	1	0	DoubleCheck=54169523,73,1
54216187	5	4	9	0	9	0	DoubleCheck=54216187,73,1
54216259	5	4	9	0	9	0	DoubleCheck=54216259,73,1
54227213	5	4	9	0	9	0	DoubleCheck=54227213,73,1
54252571	1	0	2	0	2	0	DoubleCheck=54252571,74,1
54253187	1	0	4	0	4	0	DoubleCheck=54253187,73,1
54260159	1	0	3	0	3	0	DoubleCheck=54260159,73,1
54263101	1	0	3	0	3	0	DoubleCheck=54263101,73,1
54326273	1	0	2	0	2	0	DoubleCheck=54326273,73,1
54339877	1	0	3	0	3	0	DoubleCheck=54339877,73,1
54459299	1	0	1	0	1	0	DoubleCheck=54459299,73,1
54992183	12	7	1	0	1	0	DoubleCheck=54992183,73,1
55123591	1	0	1	0	1	0	DoubleCheck=55123591,73,1
55130821	1	0	1	0	1	0	DoubleCheck=55130821,73,1
55144013	5	4	9	0	9	0	DoubleCheck=55144013,73,1
55144163	5	4	9	0	9	0	DoubleCheck=55144163,73,1
55766119	1	0	3	0	3	0	DoubleCheck=55766119,73,1
55878127	1	0	2	0	2	0	DoubleCheck=55878127,73,1
56254769	1	0	3	0	3	0	DoubleCheck=56254769,73,1
57829859	1	0	1	0	1	0	DoubleCheck=57829859,73,1
57887917	1	0	5	0	5	0	DoubleCheck=57887917,73,1
57888007	1	0	5	0	5	0	DoubleCheck=57888007,73,1
57924059	1	0	5	0	5	0	DoubleCheck=57924059,73,1
57927913	1	0	5	0	5	0	DoubleCheck=57927913,73,1
57927917	1	0	5	0	5	0	DoubleCheck=57927917,73,1
58265717	1	0	3	0	3	0	DoubleCheck=58265717,73,1
58265791	1	0	3	0	3	0	DoubleCheck=58265791,73,1
58266007	1	0	3	0	3	0	DoubleCheck=58266007,73,1
58433611	1	0	3	0	3	0	DoubleCheck=58433611,73,1
58434931	1	0	3	0	3	0	DoubleCheck=58434931,73,1
58435007	1	0	3	0	3	0	DoubleCheck=58435007,73,1
58711643	1	0	3	0	3	0	DoubleCheck=58711643,73,1
58850161	1	0	2	0	2	0	DoubleCheck=58850161,73,1
58900157	1	0	3	0	3	0	DoubleCheck=58900157,73,1
58900201	1	0	3	0	3	0	DoubleCheck=58900201,73,1
58900577	1	0	3	0	3	0	DoubleCheck=58900577,73,1
58905389	1	0	3	0	3	0	DoubleCheck=58905389,73,1
58930367	1	0	1	0	1	0	DoubleCheck=58930367,73,1
58983083	1	0	3	0	3	0	DoubleCheck=58983083,73,1
58983313	1	0	3	0	3	0	DoubleCheck=58983313,73,1
58985287	1	0	3	0	3	0	DoubleCheck=58985287,73,1
59315257	1	0	2	0	2	0	DoubleCheck=59315257,73,1
59316211	1	0	1	0	1	0	DoubleCheck=59316211,73,1
59593013	1	0	3	0	3	0	DoubleCheck=59593013,73,1
59593181	1	0	3	0	3	0	DoubleCheck=59593181,73,1
59898823	1	0	4	0	4	0	DoubleCheck=59898823,73,1

Madpoo · 2017-05-09, 15:23

Quote:

Originally Posted by LaurV

Yes, this is supposed to happen, it appears as verified, because two different programs were used, but James' parser is clever enough and identifies that only one shift value was used. I will not do a TC, because in that case a QC will still be needed (different user, blah blah). Somebody else may queue it. Thanks in advance.

It's actually a funny side effect of how the data is parsed in a PHP array where it folds (or overwrites, basically) the duplicate rows because there's nothing to differentiate the two results essentially.

The XML view of the same page shows both.
76453229

One of these days I plan to re-write how that exponent report page is done; make it more efficient.

rudi_m · 2017-05-09, 18:42

Quote:

Originally Posted by Madpoo

Here's a "top 100" list of potentially bad runs, looking at their monthly as well as total track records.

Thanks! I took all below 51M.

Mark Rose · 2017-05-10, 02:49

Quote:

Originally Posted by Madpoo

It's actually a funny side effect of how the data is parsed in a PHP array where it folds (or overwrites, basically) the duplicate rows because there's nothing to differentiate the two results essentially.

The XML view of the same page shows both.
76453229

One of these days I plan to re-write how that exponent report page is done; make it more efficient.

I would appreciate a way of getting a list of exponents, not a range, if possible.

Prime95 · 2017-05-11, 02:37

Quote:

Originally Posted by Madpoo

Here's a "top 100" list of potentially bad runs, looking at their monthly as well as total track records. I re-sorted afterwards by exponent to make things tidier (didn't do that last time...)

I took these:

51578843 3 2 4 0 4 0 DoubleCheck=51578843,73,1
51578929 3 2 4 0 4 0 DoubleCheck=51578929,73,1
51579013 3 2 4 0 4 0 DoubleCheck=51579013,73,1
51579097 3 2 4 0 4 0 DoubleCheck=51579097,73,1
51751211 1 0 1 0 1 0 DoubleCheck=51751211,73,1
51984847 1 0 3 0 3 0 DoubleCheck=51984847,73,1
52065473 1 0 1 0 1 0 DoubleCheck=52065473,73,1
52094267 1 0 3 0 3 0 DoubleCheck=52094267,73,1

ewmayer · 2017-05-13, 01:58

The quartet of bad boys ~53M copied below would be an excellent accuracy test of my 2816K FFT length in Mlucas - but precisely what does this notation mean?

Code:

exponent	Bad	Good	Unk	Sus	Solo	Mis	worktodo
43998749	1	0	1	0	1	0	DoubleCheck=43998749,72,1
...
53647547	5	4	9	0	9	0	DoubleCheck=53647547,73,1
53648423	5	4	9	0	9	0	DoubleCheck=53648423,73,1
53648893	5	4	9	0	9	0	DoubleCheck=53648893,73,1
53648981	5	4	9	0	9	0	DoubleCheck=53648981,73,1

The 4 expos at bottom are the just below my default 2816K limit - but what does '5 bad, 4 good' mean?

Mark Rose · 2017-05-13, 03:40

Quote:

Originally Posted by ewmayer

The quartet of bad boys ~53M copied below would be an excellent accuracy test of my 2816K FFT length in Mlucas - but precisely what does this notation mean?

Code:

exponent	Bad	Good	Unk	Sus	Solo	Mis	worktodo
43998749	1	0	1	0	1	0	DoubleCheck=43998749,72,1
...
53647547	5	4	9	0	9	0	DoubleCheck=53647547,73,1
53648423	5	4	9	0	9	0	DoubleCheck=53648423,73,1
53648893	5	4	9	0	9	0	DoubleCheck=53648893,73,1
53648981	5	4	9	0	9	0	DoubleCheck=53648981,73,1

The 4 expos at bottom are the just below my default 2816K limit - but what does '5 bad, 4 good' mean?

It means that machine turned in 5 bad results, 4 good results, 9 results with an unknown status, 0 that were suspect, and 0 mismatches. I haven't figured out what Solo means.

ewmayer · 2017-05-13, 04:25

Quote:

Originally Posted by Mark Rose

It means that machine turned in 5 bad results, 4 good results, 9 results with an unknown status, 0 that were suspect, and 0 mismatches. I haven't figured out what Solo means.

Ah, that makes sense, and explains why all 4 expos have those numbers the same. Assuming 0-residue-shift is OK for the DCs, I'll take those.

Aaron, do I submit the results via the manual tests page or PM them to you?

Mark Rose · 2017-05-13, 04:58

Quote:

Originally Posted by ewmayer

Ah, that makes sense, and explains why all 4 expos have those numbers the same. Assuming 0-residue-shift is OK for the DCs, I'll take those.

Aaron, do I submit the results via the manual tests page or PM them to you?

You should register them with PrimeNet either manually or using Prime95 so that no one else starts working on them. Submit how you usually submit PrimeNet assignments. I've submitted well over a thousand results this way.

2017-05-13, 01:58	#1449
ewmayer ∂²ω=0 Sep 2002 República de California 5×17×137 Posts	The quartet of bad boys ~53M copied below would be an excellent accuracy test of my 2816K FFT length in Mlucas - but precisely what does this notation mean? Code: exponent Bad Good Unk Sus Solo Mis worktodo 43998749 1 0 1 0 1 0 DoubleCheck=43998749,72,1 ... 53647547 5 4 9 0 9 0 DoubleCheck=53647547,73,1 53648423 5 4 9 0 9 0 DoubleCheck=53648423,73,1 53648893 5 4 9 0 9 0 DoubleCheck=53648893,73,1 53648981 5 4 9 0 9 0 DoubleCheck=53648981,73,1 The 4 expos at bottom are the just below my default 2816K limit - but what does '5 bad, 4 good' mean?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Double-Double Arithmetic	Mysticial	Software	52	2021-04-23 06:51
Clicking an exponent leads to 404 page	marigonzes	Information & Answers	2	2017-02-14 16:56
x.265 half the size, double the computation; so if you double again? 1/4th?	jasong	jasong	7	2015-08-17 10:56
What about double-checking TF/P-1?	137ben	PrimeNet	6	2012-03-13 04:01
Double the area, Double the volume.	Uncwilly	Puzzles	8	2006-07-03 16:02

2017-05-08, 05:47	#1443
LaurV Romulan Interpreter Jun 2011 Thailand 23·419 Posts	Yes, this is supposed to happen, it appears as verified, because two different programs were used, but James' parser is clever enough and identifies that only one shift value was used. I will not do a TC, because in that case a QC will still be needed (different user, blah blah). Somebody else may queue it. Thanks in advance.